Change search
ReferencesLink to record
Permanent link

Direct link
Improving ABR recording for diagnostic evaluation
2004 (English)Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

People with hearing disorder may have damages to their auditory pathway, which stretches from the outer ear to the auditory parts of the brain. Invasive measuring of the ABR (Auditory Brainstem Response) gives an indication of what part of the pathway that is not working properly. Because of the low signal-to-noise ratio, an averaging method is implemented in a Texas Instruments DSP (Digital Signal Processor) TMS320C6201 EVM (EValuation Module). When measuring the ABR, a stimulus in the form of a click sound or tone is put into the ear. In synchronization with the stimulus, recording of the activity in the auditory pathway with an biomedical amplifier and electrodes attached to the test subjects' head is performed. The signal-to-noise ratio for the recorded signal is low and therefore an averaging of several recordings is needed to get a visual response of the ABR. The recorded data is exported from the DSP to data files readable by MATLAB for presentation and further analysis. Implementation in the DSP is made with Texas Instruments CCS (Code Composer Studio). With the CCS program a module structure is made, making it easy to change the behavior of the DSP. A simple text based user interface is created for loading stimulus and recording data, which is done in real time. The averaging is implemented with the possibility to exclude frames containing large artifacts that corrupt the averaging. A method for detection neural events in evoked ABR with few or no averaging has been implemented and evaluated. The method uses concatenation of ABR recordings to create artificial frequencies due to relations of individual events. MATLAB and Wavelet decomposition with the Complex Morlet wavelet were used as tools. The result shows that the concatenation method doesn't work due to properties of the wavelet in use. The hypothesis requires the wavelet to have narrow time and frequency resolution that is better than can be achieved.

Place, publisher, year, edition, pages
Keyword [en]
Technology, Signal processing, ABR, Auditory Brainstem Response, Wavelet, Complex Morlet, Neural event detection, DSP, Texas, Instrument, TMS320C6201, CCS, Code Composer Studio, Concatenation
Keyword [sv]
URN: urn:nbn:se:ltu:diva-42838ISRN: LTU-EX--04/137--SELocal ID: 0cd6c73d-f134-4b7a-8880-5f6fae1068bdOAI: diva2:1016064
Subject / course
Student thesis, at least 30 credits
Educational program
Electrical Engineering, master's level
Validerat; 20101217 (root)Available from: 2016-10-04 Created: 2016-10-04Bibliographically approved

Open Access in DiVA

No full text

Search outside of DiVA

GoogleGoogle Scholar

Total: 7 hits
ReferencesLink to record
Permanent link

Direct link